gsaini

System One Models Getting Started — Development skill for Claude Code

Development community

TypeSafe's System One launch claims, each as a runnable experiment: Jev vs Claude vs keyword rules through one typed interface.

How to install System One Models Getting Started

This entry records only its repository, not the path inside it, so there is no exact command to give. Open gsaini/system-one-models-getting-started and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What System One Models Getting Started does

TypeSafe's System One launch claims, each as a runnable experiment: Jev vs Claude vs keyword rules through one typed interface. Calibration, parallel questions, Wikiracing. Offline in CI; --live for the real thing.

Alternatives in Development

Claude Code Rev — Runnable ClaudeCode source code 3k ★
Getting Started With Spawner Skills — Getting Started with Spawner Skills skill 847 ★
VibeSec Skill — to avoid getting hacked 731 ★

README

⚡ System One Models — Getting Started

**TypeSafe's launch post makes six claims about Jev, the first "System One model". This repo turns each one into an experiment you can run.** Jev, an LLM, and plain keyword rules all answer the *same typed questions* through the *same* `system_one(state, questions)` call, and code measures the rest: latency, tokens, cost, calibration, and whether the answers are right.

[![CI](https://github.com/gsaini/system-one-models-getting-started/actions/workflows/ci.yml/badge.svg)](https://github.com/gsaini/system-one-models-getting-started/actions/workflows/ci.yml) ![Python](https://img.shields.io/badge/Python-3.11%2B-3776AB?style=for-the-badge&logo=python&logoColor=white) ![TypeSafe SDK](https://img.shields.io/badge/typesafe--sdk-0.7.4-E551BA?style=for-the-badge) ![Adapter](https://img.shields.io/badge/system--one--adapter-0.2-6B21A8?style=for-the-badge) ![Claude](https://img.shields.io/badge/LLM%20arm-Claude%20Opus%205.5-D97757?style=for-the-badge&logo=anthropic&logoColor=white) ![Offline](https://img.shields.io/badge/Tests%20%26%20baselines-offline%2C%20free-16A34A?style=for-the-badge) ![License](https://img.shields.io/badge/License-MIT-EAB308?style=for-the-badge)

An independent learning project, not affiliated with TypeSafe or Anthropic. The post is dated 2026-10-13 and this was built the next day; prices, limits, and the SDKs are changing weekly, so re-check anything you rely on.

The post in one paragraph

[*Introducing System One Models & Jev*](https://typesafe.ai/blog/introducing-system-one-models-and-jev) (Diogo Almeida, TypeSafe, 2026-10-13) describes a new class of model that **doesn't generate text**. You send a *state* (text or JSON) and typed *questions*; you get back typed answers with probabilities: a **Choice** from options you listed, a **Score** on levels you described, a **Noul** (P(yes) for a statement). The post calls it *"a frontier-intelligence function call: unstructured state in, typed probabilisti